largely accounted for by a lower birth rate in Madras—34.9 compared to 41.7 for all of India, according to census estimates. Moreover, if the family planning program of the government of India has any significant effect on fertility rates, the gap between Madras and other states may increase, since a larger proportion of couples have adopted family planning methods in Madras in recent years than in all but three other states. According to a recent study, 18 percent of Madras couples are "protected" against 12.1 percent for the entire country (55). Both within and between states, the population of ethnic groups has become an important element in the struggle for political power and in influencing public policy decisions. As Indian politicians during the past two decades have become increasingly sophisticated about the relationships between ethnicity, population size, and voting power, they have also become increasingly concerned with changes in the composition and redistribution of populations and with the ways in which such data is recorded in the census. On these matters, Indian politicians are considerably more advanced than are political leaders in much of Africa, where ethnic groups are often less cohesive politically, where the absence of representative institutions reduces the importance of numbers, and where census-collecting institutions are not yet developed. In the future, we can expect population changes in Africa, especially differential fertility rates and migration rates, to become, as in India, sources of political conflict. Political Effects of Migration In traditional societies, redistributions in population groups are generally due to differential fertility and mortality rates. In all modern societies migration is the major element in the geographic redistribution of populations. As societies modernize and become more urban, internal migration becomes more important and the geographical redistribution of populations increases. In the United States regional population shifts either into cities or into states have been largely a consequence of migration. In India redistribution of populations between states, at least before 1951, was due primarily to differential natural increases. Zachariah estimates, however, that, since 1951, changes in the proportion of the Indian population living in each of the states are due more to internal migration than to differential natural increases (56, pp. 93-106, especially Table 3; 41). To what extent is population growth a factor in migration? As population increases, do people move from the densely populated areas of the country to the less crowded regions? On the contrary. In the last two decades in the United States, for example, as urban density has been increasing, half of the 3,000 rural counties have continued to lose population to the crowded cities. Similar trends are seen in many less developed countries where one or two have expressed their concern over the long-term position of Madras in the national Parliament, in which representation is determined by population. The difference in growth rate seems cuts both ways. less important.